PROFESSOR STOKES ON THE CHANGE OF REFRANGIBILITY OF LIGHT. 549 
to dynamical principles to suppose that any sueh causes should be adequate to 
account for the production of vibrations of one period from vibrations of another. 
All believers, I suppose, in the undulatory theory of light are agreed in regard- 
ing the production of light in the first instance as due to vibratory movements 
among the ultimate molecules of the self-luminous body. Now in the phenomenon 
of internal dispersion, the sensitive body, so long as it is under the influence of the 
active light, behaves as if it were self-luminous. Nothing then seems more natural 
than to suppose that the incident vibrations of the luminiferous ether produce vibra- 
tory movements among the ultimate molecules of sensitive substances, and that the 
molecules in turn, swinging on their own account, produce vibrations in the lumini- 
ferous ether, and thus cause the sensation of light. The periodic times of these 
vibrations depend upon the periods in which the molecules are disposed to swing, 
not upon the periodic time of the incident vibrations. 
227. But in the very outset of this theory an objection will probably be urged, that 
it is quite as much contrary to dynamical principles to suppose the periodic time of 
the ethereal vibrations capable of being changed through the intervention of ponder- 
able molecules as without any sueh machinery. The answer to this objection is, that 
such a notion depends altogether on the applicability of a certain dynamical principle 
relating to indefinitely small motions, and that we have no right to regard the mole- 
cular vibrations as indefinitely small. The excursions of the atoms may be, and 
doubtless are, excessively small compared with the length of a wave of light ; but 
it by no means follows that they are excessively small compared with the linear 
dimensions of a complex molecule. It is well known that chemical changes take 
plaee under the influence of light, especially the more refrangible rays, which would 
not otherwise happen. In such cases it is plain that the molecular disturbances 
must not be regarded as indefinitely small. But vibrations may very well take place 
which do not go to the length of complete disruption, and yet which ought by no 
means to be regarded as indefinitely small. Furthermore, it is to be observed that if 
in the cases of indefinitely small molecular displacements the forces of restitution be 
not proportional to the displacements, the principle above alluded to will not be ap- 
plieable however small the disturbance may be ; and if in the expressions for the 
forces of restitution the terms depending on first powers of the displacements (sup- 
posed finite), though not absolutely null, be very small, the principle will not apply 
unless the molecular excursions be extremely small indeed. In consequence of the 
necessity of introducing forces not proportional to the displacements, it would be 
very difficult to calculate the motion, even were we acquainted with all the circum- 
stanees of the case, whereas we are quite in the dark respecting the actual data of 
the problem. But certainly we cannot affirm that in the disturbance communicated 
back again to the luminiferous ether none but periodic vibrations would be produced, 
having the same period as the incident vibrations. Rather, it seems evident tliat a 
sort of irregular motion must be produced in the molecules, periodic only in the 
